I've been trying to figure out how to use Fitt's Law properly to
increase the targetability of a dense tree structure, as displayed on
a browser page.
The information in the hierarchy is three levels, call them Category,
Subcategory, and Items.
There are about 5 Categories, each of which has 3-10 Subcategories and
each subcategory has 1-15 Items.
Displaying these data in a conventional tree leads to a display that
is a grand pain to target on. The left side is nice and large and
clear, but the right column of Items is dense and packed so that
targeting becomes quite slow.
Ideally I'd replace the tree with something like a pie menu or other
progressive display, except that the users are infrequent visitors.
We cannot expect them to learn or remember the taxonomy. The display
needs to support scanning.
In addition, a significant fraction of the users know the NAME of the
thing they want, but again do not know or remember the categories.
Thus I want to support conventional "find" functionality (ctrl-F).
of the Items group when the user's mouse got nearby, but that seems to
confuse people as their target appears to "jump" and may not be where
they're tracking to. Expanding the font itself (by bolding or changing
size) also has bad effects in that it can cause word wrap, meaning the
target is suddenly very far away from where the person is tracking.
Can anyone suggest a way I can meet the two constraints (support
scanning; support find-ability by name) while improving the
time-to-target problem for people who are actually using the taxonomic